1899] DEVELOPMENT OF THE MICROSPORANGIUM 345 
mw longer any doubt but that the tapetum is differentiated from 
the wall, rather than from the archesporium, as a hasty inspec- 
tion would seem to indicate. 
The wall of the anther at this stage is composed of three, or 
mely four, layers of cells (fig. 26). The outermost of these 
ayers remains almost unchanged until the anther is mature, and 
‘Sindeed the true epidermis. The other two or three layers 
lave some of their cells arranged in more or less distinct radial 
Ws, Suggesting, as in the first case, that each row is the deriva- 
tive of one hypodermal cell. The greater portion of the wall, 
lowever, is formed, as in Convallaria, from the cells of either 
side of the archesporium, and these are not necessarily deriva- 
tives of a hypodermal cell. Indeed, so far as could be deter- 
tined, the growth was brought about exactly as in Convallaria. 
ing the maturation of the anther the behavior of the cells is 
‘mal, The third layer undergoes disintegration, as does also 
‘he fourth, Which is the tapetum. The epidermis remains nor- 
mi,while at the same time the second layer becomes thicker 
walled than the rest, acquires spiral or reticulated thickenings, 
‘itis indeed a true endothecial layer ( fig. 27). 
\ Campbell considers the anther of Naias to be a so-called 
/Gulome” Structure, in which are early differentiated plerome 
be periblem, the upper cell of the plerome cylinder becoming 
~* Mtchesporium, This in itself does not preclude a process 
ad to that described above for Potamogeton ; although 
. Pbell himself js quite certain that the origin of the arche- 
& i Naias is not traceable to a single cell. He found 
tee of only two layers besides the pigs 
A tacin ree. In Zannichellia the same difficulty was fo 
‘ym, the development, but this was probably because the 
thy are “ay Not counted as a wall layer. The satu 
et cat ere also at first scarcely distinguishable from ae a a 
®hesporis In this plant there are three layers surroun ing 
py Mall of which finally become disintegrated. 
plete disintegration of the tapetal cells in Potamogeton 
o Soincident with the divisions of the pollen-mother-cell. 
ane COm 
Salm 
